Back to search

BIA-Brukerstyrt innovasjonsarena

Decision support system for professional real-estate price estimation

Alternative title: Beslutningsstøtteverktøy for profesjonell eiendomsprisestimering

Awarded: NOK 9.7 mill.

Project Manager:

Project Number:

332166

Project Period:

2022 - 2024

Funding received from:

Organisation:

Location:

This project will employ digitalization and artificial intelligence to develop a decision-support system for real estate professionals. The system will improve price estimates, support experts’ assessment processes, and improve renovation of homes. The objective of the project is to develop tools that will: 1) Improve the accuracy and speed of predicting the selling price of a home. 2) Determine the price contributions of specific features and amenities (apartment floor, size, parking, floorplan, etc.). 3) Determine how the price of a home can increase through renovations or home improvements. The project will create prototypes of a decision-support system and price estimation algorithms. Key inputs to this work will be previous assessments of objects, novel data about objects such as floor plans and landscape data, and captured expert knowledge on price estimation and renovation. The goal is to combine this expert knowledge with statistical data from large datasets of previous transactions and house features. To do this, the prototyped decision-support system will provide experts with task-relevant information and allow operators to inspect, modify and override automated decision logic in the system. This will improve the validity range of valuations as well as valuation speed and accuracy.

The real estate industry is the largest asset class in Norway. In 2020 alone 125,000 homes were sold representing almost NOK 500 billion in transaction value. Nevertheless, while there is an increasing number of industries that are digitizing and automating processes, the typical process of buying or selling a home remains unchanged. To address this, Solgt.no has launched an Instant Buyer (or iBuyer) service, wherein the company purchases residential properties directly from private sellers, to eventually re-sell them providing, liquidity, reduced risk and convenience to the customer. A critical aspect of the process is timely and precise valuation as an underpricing will reduce the chance of a profitable purchase, while an overpricing will reduce the profit (or potentially incur a loss). Similarly, accurate and timely dwelling valuation is of critical importance in real estate development. The earnings potential of a property development is determined by dwelling pricing. Incorrect pricing of apartments is the single biggest reason behind loss-making projects. Furthermore, the number of potential development prospects is limited by the available time to evaluate each prospective development. Motivated by these applications, we have identified a significant opportunity to create a Decision Support System (DSS) for real-estate price estimation by firstly, digitizing and automating parts of the valuation process of domain experts, secondly, engineering new valuation features from publicly available data, and thirdly, incorporating unique datasets of customer and developer data from the project participants. The envisioned DSS will deliver dwelling price estimates at a fraction of the time of that of current work processes, and with a step change improvement in accuracy.

Funding scheme:

BIA-Brukerstyrt innovasjonsarena